In this proposal, we build on our previous work using Positron Emission Tomography (PET) to contribute to the management of sarcoma patients. Our proposed studies in sarcoma patients also serve as a model for PET applications in other types of cancer. Using [F-18] fluorodeoxyglucose (FDG) and PET/CT, we will perform prospective imaging studies on patients to determine their tumor metabolism for identification of treatment response, and the ability of these measures to predict outcome. This work will continue our evaluation of the ability of FDG PET derived tumor measures to assess and predict time to local tumor recurrence, metastasis and death using rigorous statistical methods based on standard epidemiology concepts. Imaging data as surrogate endpoints for clinical outcome will be tested with hypothesis driven studies for validation with clinical outcomes data. The first specific aim will address characterization of tumor response to neoadjuvant chemotherapy in patients with soft tissue sarcomas. This image derived response measure will be validated by correlation with patient outcome to determine the ability of FDG PET/CT imaging to be used as a surrogate endpoint for treatment effectiveness in this group of patients. In the second specific aim, FDG PET imaging will be used in a similar manner to characterize treatment response in the pediatric and young adult population of patients with osteosarcoma and Ewings Sarcoma Family of Tumors (ESFT) which includes peripheral neuroectodermal tumors. As in the previous aim, we will evaluate the ability of the FDG PET image derived measure of tumor response to predict patient outcome after neoadjuvant chemotherapy. In a third specific aim, we will continue our previous work in quantitating measures of tumor heterogeneity in FDG metabolism using our statistical image analysis methods for complex tumor FDG uptake distribution. These data will be analyzed for their ability to predict tumor response, and patient overall survival. This new image analysis algorithm will be applied to generate a semi- quantitative and spatial assessment of the tumor surface with statistical methods. The focus of these studies are to exploit the quantitative data and high level of spatial sensitivity in FDG PET images to characterize tumor metabolism and to continue to develop image analysis methods that provide relevant data and insights on tumor biologic aggressiveness. [unreadable] [unreadable] [unreadable]